any good tutorials on best practices for collision boxes? Sounds like it's actually pretty confusing for a few users.
BBaller1337
mudmask - I don't think we have an in depth tutorial on collision boxes in C2 itself aside from what's in the tutorial thread linked in my signature, but it's actually pretty simple. First, the recommended method for doing something like using your entire character with platform behavior is just to pin the scml object to a dummy sprite (which is how it's usually done for regular sprites in C2 as well). If you want separate animate-able collision boxes, this works using the old method of importing. Import your project, and Construct 2 will create invisible sprite objects with the names of your collision boxes. Simply check for collisions the way you would on any sprite in C2, and since everything is in a group, it should automatically pick the correct instance of your scml object for further actions.
Unfortunately, there are no plans to add collision boxes directly into the new direct drawing mode, as it's a bit of a slippery slope, where updates to C2's collision detection methods will always necessitate changes to the plugin's as well, and I would expect it to get messy and potentially cause performance issues. For anyone curious, there is actually a hybrid way to import now, which is a bit tedious, but I will explain it in case anyone wants to use direct drawing mode with collision boxes, or even with collideable sprites. First, save your project normally, and then save a spritesheeted project. Import your non-spritesheeted project the old way. Follow the old to new style project conversion steps, but this time only delete everything except collision boxes, and other sprites you want to test for collisions. This should self draw the animation, but also animate the separate sprites and collision boxes you kept. Keep in mind, the zorder will not place sprites correctly inside the self-drawing scml object, so it's probably only useful for making invisible sprite objects for collision detection.
EDIT: 
lucid sent the e-mail
Thank you. I will look into it and respond as soon as I get a chance.
Looking for a method where the recolor is purely visual so it should reuse all the same other data, but without me having to update changes constantly, if possible. Let me know if you have any ideas.
The best method is to recolor the individual images, and then use character maps to share the animation data and to switch colors on the fly in C2. A tutorial for character maps in C2 is in the main tutorial page linked in my signature, and more information about character maps in Spriter can be found under Help in Spriter.
I remember seeing a post or something for how to preload animations (or something like it) to help the game run faster, but I can't find it. Anyone know what I'm talking about or just have some tips? I'm really struggling on optimization here.
You can try the new method of importing, which has faster performance:
My other question is if it's better to have a spriter object have a separate object for each animation (so in C2 an idle animation would be one sprite object and a run animation would be another one), or to have one object for all animations.
It's best to have one scml file (and one scml object in C2) for each character (or machine, or any individual animated object).
lucid - How do we spawn object by image point using this new plugin? Is there a new way? In the past plugin I made an image point on the separate sprites in order to spawn objects for example gun sprite on a spriter object to spawn bullets. Thanks.
—, you would have to use Spriter's action point feature. which let's you place named points, and optionally attach them to bones and animate them. The main tradeoff for using the new self-drawing mode is that you don't get the advantages of having separate C2 objects.